Control of time-delay systems through modified Smith predictor using sliding mode controller

Smith-Predictor (SP) is used as a dead-time compensating tool for chemical processes as they are Integrating-Plus-Dead-Time (IPDT), pure integrating-type process, and higher-order-integrating-process. The performances of these processes can be improved by modifying SP. The present work deals with the control of dead-time by integrating the first-order-plus-dead-time (IFOPDT) or first-order-plus-dead-time process (FOPDT), and a modified sliding-mode-controller (SMC) in the proposed modified-Smith-predictor (MSP) structure. Here, a modified novel SMC discontinuous tuning-parameter is derived analytically to satisfy performance criteria such as speed adjustment, reducing overshoot and chattering effects. Furthermore, the MSP-SMC structure is implemented in six different processes. This is done to counteract time-delay and constant load-disturbance for integrating processes and a nonlinear system separately. Also, the proposed work is implemented in the real-time lab setup of a neutralisation process. Furthermore, the robustness and invariance properties of the SMC have been analysed against parametric uncertainty. The optimality of the proposed controller parameters is tested by giving uncertainty in the dead time of the process parameters. The proposed design is also implemented in the non-linear and Multi-Input-Multi-Output processes. The performance metrics, IAEs, of 6 case studies have been obtained using MATLAB and the peak overshoots are compared with similar kinds of works.The discontinuous control law of SMC_MSP tuning equation KD has been modified.Modified novel KD equation is derived ANALYTICALLY to satisfy the performance criteria and implemented in the modified Smith predictor structure to counteract the dead time and constant load disturbanceSet point filter implemented eliminates the peak overshoot considerably.Set point tracking, disturbance rejection and robustness have been analysed.IFOPDT, ISOPDT, IHOPDT, Nonlinear CSTR and MIMO processes were simulated.